Age estimation through sternal fusion and costal cartilage ossification using MSCT in a Croatian population: model development and application.

Abstract

This study aimed to test age-related changes in sternal fusion and sternal-rib cartilage ossification on multi-slice computed tomography (MSCT) images of the Croatian population. The additional aim was to develop models to estimate age and provide an interface for the model's application and validation. This retrospective study was conducted on 144 MSCT images of the sternal region, and the developed models were tested on 36 MSCT images. We scored manubrium-mesosternal joint fusion (FM), xiphoid process and mesosternum fusion (FX), ossification of the first costal cartilage (OF), ossification of the second to seventh costal cartilages at the rib ends (OR), and ossification of the second to seventh costal cartilages at the sternal ends (OS). All sternal-rib cartilage ossification phases and sternal body and xiphoid process fusion scores showed statistically significant age differences (P < 0.001), except manubrium-mesosternal joint fusion. The final model that combined regression and classification using FM, FX, OR, OS, and sex obtained a 95% prediction interval (PI) coverage of 94.46% on the cross-validation (cv) and 91.67% on the test set with an average PI width of 42.29 and 42.95 years respectively. We also developed a Python Flask app called CroSterna: Age estimation from sternal fusion and rib ossification in the Croatian population ( https://crosterna.onrender.com/ ) to facilitate the estimation for professionals.